Improved Spray Paint Thickness Calculation From Simulated Droplets Using Density Estimation

Abstract

Advancements in the simulation of electrostatic spray painting make it possible to evaluate the quality and efficiency of programs for industrial paint robots during Off-Line Programming (OLP). Simulation of the spray paint deposition process is very complex and requires physical simulation of the airflow, electric fields, breakup of paint into droplets, and tracking of these droplets until they evaporate or impact on a surface. The information from the simulated droplet impacts is then used to estimate the paint film thickness. The current common way of measuring paint thickness on complex geometrical shapes is to use histogram based methods. These methods are easy to implement but are dependent on good quality meshes. In this paper, we show that using kernel density estimation not only gives better estimates but it also is not dependent on mesh quality. We also extend the method using a multivariate bandwidth adapted using estimated gradients of the thickness. To show the advantages of the proposed method, all three methods are compared on a test case and with real thickness measurements from an industrial case study using a complex automotive part.